Restriction Enzymes 119
scissile phosphodiester bond, rather than with an involvement of a
covalent enzyme intermediate (115). It remains to be established whether
this is true for other restriction enzymes (see note added in proof at end
of chapter). Since some nucleases show retention of configuration at
phophorous, and others show inversion of the configuration, it might well
be that not all restriction endonucleases utilize the same mechanism.
2.4. Specificity
The specificity of restriction endonucleases for their recognition
sites is very high. Under optimum reaction conditions, even sites that
differ in only one base pair from the canonical site are not cleaved,
unless large enzyme concentrations are used or the reaction is allowed
to proceed for sufficient time. For example, a standard pUC8 DNA
digestion carried out with micromolar concentrations of EcoRI leads
to cleavage within seconds not only at the canonical sequence -GAATrC-,
but after several hours of incubation also at two other sites, -TAATTC-
and -GAGTTC-; two other sequences in pUC8 that also differ in only
one base pair from the canonical site, -TAATTC- and -GAATAC-,
however, are refractory to cleavage (Pingoud and Alves, unpublished).
Similarly, EcoRV cleaves not only its recognition sequence, -GATATC-,
but also, at very small rates, several alternative DNA sequences, e.g.,
in pAT153 -GAAATC- (116) and -GTTATC- (117). A detailed study
using a series of nine tetradekadeoxynucleotides that contain all pos-
sible single base pair substitutions of the EcoRI recognition site has
shown that, depending on the position within the sequence and the
kind of substitution, the oligonucleotides are cleaved more or less
readily, with rates varying from 1/1000 to <1/106 of the rate with
which the canonical sequence is cleaved (114) (see note added in proof
at end of chapter).
Considerable progress has been made in recent years in understand-
ing the molecular basis of the specificity of restriction enzymes, mainly
because of chemical modification studies, in which the structural ele-
ments of the DNA substrates needed for efficient cleavage were identi-
fied (ref. 5-64 in [87]). In spite of the intrinsic problem of differentiating
between direct and indirect effects caused by the chemical modifica-
tion of the oligodeoxynucleotide on the acceptance as a substrate by
the enzyme, it is clear from these studies that a lot but not all of the
structural information present in the sequence of bases that constitute